Although it has been known for some time that certain retroviruses cause leukemia in experimental animals, the mechanisms by which these retroviruses cause leukemia is not well understood. Clonal hemopoietic cell culture appears to provide an excellent in vitro experimental system in which to study the effects of such viruses directly on hemopoietic progenitors. Our approach is to isolate individual hemopoietic progenitors in vitro and to use these cells as targets for retroviral gene transfer. Subsequent culture of the infected cells permits in vitro analysis of the expression and effects of the viral gene as the primitive cell proliferates and executes its developmental program. We have infected individual progenitors from blast cell colonies with Harvey sarcoma virus. Not all the cells in colonies infected by Harvey sarcoma virus were positive for the ras protein as determined by immunofluouresence. This raised the possibility that the v-ras was preferentially expressed at some stages of differentiation. We propose two approaches for the study of variable gene expression, in situ hybridization and an immunoperoxidase method, which allow detection of viral gene expression and preserve sufficient cell morphology for cell identification. In our previous study, the progeny of the infected adult pluripotent hemopoietic cells did not demonstrate a transformed phenotype even though they expressed the v-ras gene. We recently developed a new progenitor assay using fetal liver cells that provides a source of committed erythroid and mast cell progenitors. This population appears to be susceptible to transformation by Harvey sarcoma virus. Transformation will be monitored by changes in growth rate, maturity, sensitivity to growth factors or self renewal ability. We will develop a serum- free culture system for the delineation cf the growth factor requirements of purified fetal liver pluripotent and committed pro- genitors. The serum-free system will be applied first to fetal liver blast cell colonies then to progenitors purified by elutriation and cell sorting. Finally, we will test for tumorigenicity in vivo using isolated progenitors infected with helper-free stocks. The transfer of specific gene constructs into multipotent cells and committed progenitors and the analysis of expression in individual cells should increase our ability to study the genetic elements responsible for lineage-specific gene expression in a controlled environment containing purified or recombinant growth factors.